Font Size: a A A

Study Of Visual Network Management System Based On The Content Type

Posted on:2008-07-19Degree:MasterType:Thesis
Country:ChinaCandidate:X Z XiaFull Text:PDF
GTID:2178360212993683Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
With the development of computer network and the extension of its application scope, network plays an important role in people's daily work and life. However, due to the increase in complexity and heterogeneity of the network, more and more viruses spread via network and the malicious behaviors caused by people on purpose increase, which seriously affect capacity of network and disturb its regularity. In this case, how to develop an efficient network management system for network manager becomes an exigent research issue.Because of above reasons, this thesis designs a visual system of network management based on content type running in Linux system. This visual system of network management, built upon open resource project, was developed on the platform of Zope content type management architecture, mostly programmed with Python, combined with Ajax, XML, CSS and RRDtool to enrich the performance of Web interactive pages. The entire system is divided into five layers of object, service, data, logic and presentation, and five function modules of discovery, monitor, collection, alert and display, with clear arrangement and rational structures. Content type database ZODB and relational database MySQL are used as background date storage in the system, which ensure stable and efficient management performance for the system under the environment that a lot of different types of data need to be stored.The visual system of network management designed and realized by this thesis has three major innovations.Firstly, it introduces the currently prevalent thinking of content management into design of network management system, and abstracts network equipments, network services, network topology and the various views of properties to be different content type. It uses the development platform of open resource content management, based on the architecture of Web network management, which makes the system economical practical, safe, reliable, stable and efficient.Secondly, it designs a novel algorithm of multi-area and multi-subnet Ethernet physical topology discovery. This algorithm avoids difficulty brought by variety of network products and variance of two-layer topology discovery protocol. It can automatically discover exact network topology of two-layer and three-layer only by using universal SNMP. This algorithm has a character that can not only discover the management devices (e.g. routers and switches), but also the dumb devices (e.g. hosts. hubs and non-managed switches).Thirdly, it efficiently samples on the boundary high-speed traffic of campus network by using NetFlow technique, analyzes ingredient of campus network boundary traffic's address, and designs a 3D modeling tactic based on segment of address-A. Using this modeling tactic to visual render the traffic of campus network boundary, provide an intuitive method to observe the network by dint of date flow in 3D space. It can display all attributes of network traffic more clearly.Presently, to deal with the difficulty brought by inflated network, diverse network equipments and applications, model of network management system is progressing continuously, management technology is diverse constantly. This thesis has tentatively explored the architecture and means of network management and tries to do some useful work to improve campus network management.
Keywords/Search Tags:network management, content type, topology discovery, network measurement, visualization
PDF Full Text Request
Related items